SAR AND QSAR MODELING OF ALGICIDAL COMPOUNDS BASED ON PHYSICOCHEMICAL DESCRIPTORS

2025

JUNE

NEW RENAISSANCE

INTERNATIONAL SCIENTIFIC AND PRACTICAL CONFERENCE

VOLUME 2

|

ISSUE 6

445

UDK: 632.952:547.2:544.4:004.94

SAR AND QSAR MODELING OF ALGICIDAL COMPOUNDS BASED ON

PHYSICOCHEMICAL DESCRIPTORS

Pardayev Ulug‘bek Xayrullo o‘g‘li

E-mail:

pardayevulugbek125@gmail.com

A student of the Chemistry program at the Faculty of Natural Sciences, Uzbekistan-Finland

Pedagogical Institute.

Akramova Yulduz Dostonbek qizi

E-mail:

yakramova22@gmail.com

A student of the Chemistry program at the Faculty of Natural Sciences, Uzbekistan-Finland

Pedagogical Institute.

Majidova Gulhayo Abdumalik qizi

E-mail:

gulxayomajidovs@gmail.com

A student of the Chemistry program at the Faculty of Natural Sciences, Uzbekistan-Finland

Pedagogical Institute.

Xolmirzayev Mehroj Murodullayevich

Assistant Lecturer at the Department of Chemistry, Faculty of Natural Sciences, Uzbekistan-

Finland Pedagogical Institute.

https://doi.org/10.5281/zenodo.15731673

Abstract.

This study presents a comprehensive structure

–

activity relationship (SAR) and

quantitative structure

–

activity relationship (QSAR) analysis of organic compounds with algicidal

properties. By employing a dataset of structurally diverse molecules, we evaluated the predictive
power of key physicochemical descriptors, including boiling point, melting point, vapor pressure,
logP, and molecular weight. Statistical models were constructed using multiple linear regression
and validated through cross-validation techniques to assess their accuracy and robustness. The
results indicate that certain descriptors, particularly logP and vapor pressure, show a strong
correlation with algicidal efficacy. The proposed models provide a valuable framework for the
rational design and pre-screening of environmentally safe algicides, reducing the need for labor-
intensive bioassays.

Keywords:

algicidal compounds, physicochemical descriptors, SAR, QSAR, predictive

modeling, organic pesticides, environmental toxicity.

Introduction:

Harmful algal blooms (HABs) have become an increasing concern in

aquatic ecosystems, causing ecological imbalance, biodiversity loss, and serious risks to public
health and water quality. As conventional chemical treatments often lead to environmental toxicity
and non-selective bioaccumulation, the demand for safer, more targeted algicidal agents has grown
significantly. Among the promising alternatives are organic compounds with tunable structures
and physicochemical properties that can be optimized for algicidal selectivity and environmental
compatibility.

Recent advancements in cheminformatics have enabled the use of structure

–

activity

relationship (SAR) and quantitative structure

–

activity relationship (QSAR) modeling to predict

biological activity based on molecular descriptors. These models reduce the reliance on time-

2025

JUNE

NEW RENAISSANCE

INTERNATIONAL SCIENTIFIC AND PRACTICAL CONFERENCE

VOLUME 2

|

ISSUE 6

446

consuming bioassays by identifying key structural and physicochemical factors

—

such as boiling

point, vapor pressure, lipophilicity (logP), molecular weight, and hydrogen bonding capacity

—

that influence algicidal potential.

Despite progress in insecticide and herbicide QSAR modeling, limited research has

addressed the modeling of algicidal compounds using physicochemical descriptors. This study
aims to fill that gap by developing SAR and QSAR models for a diverse set of organic molecules
with known algicidal activity. By analyzing the relationships between molecular features and
bioefficacy, this work contributes to the rational design of novel, environmentally friendly
algicides.

Literature review

: The increasing prevalence of harmful algal blooms (HABs) has

prompted a growing div of research into effective and environmentally safe algicidal agents.
Traditional algicides such as copper sulfate and synthetic herbicides have demonstrated strong
algal control, but they often cause adverse environmental effects, including toxicity to non-target
organisms and bioaccumulation in aquatic food webs (Anderson et al., 2012).

In recent years, organic compounds with selective algicidal properties have received

considerable attention due to their structural diversity and biodegradability. Several studies have
explored the relationship between molecular properties and algicidal activity. For example, Zhao
et al. (2017) demonstrated that hydrophobicity (logP) and vapor pressure significantly influence a

compound’s ability to penetrate algal cell membranes. Similarly, Wang et al. (2020) showed that

compounds with moderate volatility and lower water solubility exhibited higher selectivity against

Microcystis aeruginosa

.

Structure

–

activity relationship (SAR) and quantitative structure

–

activity relationship

(QSAR) modeling have emerged as powerful tools for predicting the bioactivity of chemical
compounds based on molecular descriptors. These computational approaches have been
successfully applied in pharmaceutical and pesticide research, enabling high-throughput screening
and rational compound design (Tropsha, 2010; Cherkasov et al., 2014). However, their application
in algicide modeling remains relatively limited, with only a few studies developing predictive
QSAR models specific to algal toxicity (Yang et al., 2019).

Key physicochemical descriptors used in previous QSAR studies include boiling point,

melting point, molecular weight, topological polar surface area (TPSA), and hydrogen bond
donors/acceptors. These parameters influence compound solubility, transport, and reactivity,
which are critical to algicidal function (Karelson et al., 1996). Advanced software tools such as
PaDEL-Descriptor and the OECD QSAR Toolbox have further streamlined the generation and
analysis of such descriptors for model building.

Despite the progress made, there is still a need for robust, validated models that can

accurately predict algicidal potential from simple molecular input. This study addresses this gap
by constructing SAR and QSAR models based on experimentally tested algicides and a focused
set of physicochemical properties.

Methodology:

A curated dataset of 40 organic compounds with documented algicidal

activity was compiled from peer-reviewed literature and chemical databases such as PubChem and
ChemSpider. The compounds were selected based on availability of experimental algicidal
effic

acy data (e.g., EC₅₀, LC₅₀, or inhibition rate) against model algae species such as

Microcystis

2025

JUNE

NEW RENAISSANCE

INTERNATIONAL SCIENTIFIC AND PRACTICAL CONFERENCE

VOLUME 2

|

ISSUE 6

447

aeruginosa

and

Chlorella vulgaris

. Physicochemical descriptors for each compound were

calculated using PaDEL-Descriptor (Yap, 2011) and verified with the OECD QSAR Toolbox. The
selected descriptors included boiling point, melting point, vapor pressure, logP (octanol

–

water

partition coefficient), molecular weight, hydrogen bond donors/acceptors, and topological polar
surface area (TPSA). All values were normalized to reduce scale bias. Structure

–

Activity

Relationship (SAR) analysis was performed by grouping compounds based on shared functional
groups and structural features (e.g., aromatic rings, halogen substitutions, heterocycles) and
comparing their relative algicidal performance. Heatmaps and clustering analysis were used to
visually identify structure-dependent activity trends. Multiple linear regression (MLR) and partial
least squares regression (PLSR) techniques were used to construct quantitative models. Predictor
variables were selected using stepwise regression and Variance Inflation Factor (VIF) analysis to
eliminate multicollinearity. The response variable was the normalized algicidal activity value. The
models were evaluated using internal validation (leave-one-out cross-validation, LOO-CV) and
external validation (test/train split, 80:20). Performance metrics included the coefficient of

determination (R²), root mean square error (RMSE), mean absolute error (MAE), and predictive
R² (Q²). All data preprocessing, modeling, and statistical analysis were conducted using Python

(scikit-learn, pandas, numpy) and IBM SPSS Statistics v27. Visualization of correlation matrices
and model diagnostics was performed using matplotlib and seaborn libraries.

Results:

The structure

–

activity relationship (SAR) analysis revealed that compounds

containing halogenated aromatic rings (e.g., chloro- or bromo-substituted phenyl groups) exhibited
consistently higher algicidal activity. Heterocyclic structures such as pyrroles and thiazoles also
demonstrated enhanced activity compared to aliphatic analogs. Conversely, compounds with polar
substituents (e.g., hydroxyl or carboxyl groups) tended to show reduced efficacy, likely due to
decreased membrane permeability.

Hierarchical clustering analysis grouped the compounds into three major activity classes

—

high, moderate, and low

—

based on structural similarity and observed bioactivity values. This

qualitative assessment provided the basis for selecting descriptor sets for quantitative modeling.
(Model 1,2 and Table 1)

Table 1: Structural Features and Their Influence on Algicidal Activity:

№

Structural Feature

Example Groups

Observed Effect on

Activity

1 Halogenated Aromatic Rings

–

Cl,

–

Br substituted phenyl

rings

Strongly increased

activity

2

Heterocyclic Structures

Pyrrole, Thiazole

Increased activity

compared to aliphatics

3

Polar Substituents

–

OH,

–

COOH

Decreased activity

4

Aliphatic Chains

Saturated hydrocarbon groups

Lower activity

(baseline/reference)

5

Structural Similarity

Grouping (Clustering)

Clustered as High, Moderate,

and Low

Used for descriptor

selection

2025

JUNE

NEW RENAISSANCE

INTERNATIONAL SCIENTIFIC AND PRACTICAL CONFERENCE

VOLUME 2

|

ISSUE 6

448

Model 1 explores the relationship between algicidal activity, lipophilicity (logP), and

volatility (vapor pressure). The surface plot shows a clear trend: compounds with higher logP
values and lower vapor pressure demonstrate greater predicted algicidal activity. This suggests
that hydrophobicity enhances membrane permeability

,

allowing compounds to more effectively

penetrate algal cells, while lower volatility helps maintain sufficient contact time with the target
organisms. The surface is smooth and exhibits a strong gradient along both axes, supporting the
significance of both descriptors in activity prediction.

2025

JUNE

NEW RENAISSANCE

INTERNATIONAL SCIENTIFIC AND PRACTICAL CONFERENCE

VOLUME 2

|

ISSUE 6

449

Model 2 investigates the effect of lipophilicity (logP) and topological polar surface area

(TPSA) on algicidal activity. The plot reveals that high logP

coupled with low TPSA

leads to

higher predicted activity. This confirms that polar surface area negatively influences bioactivity,
possibly due to reduced cellular uptake of highly polar molecules. While the effect of logP remains
positive, increasing TPSA appears to systematically lower the predicted efficacy, reflecting its role
in decreasing membrane permeability and increasing hydrophilicity.

Table 2: Comparison of QSAR Models Based on Descriptor Pairs:

№

Model

Descriptors

Used

Positive

Influence

Negative

Influence

Optimal

Range

Observed

1

Model 1

logP, Vapor

Pressure

logP

Vapor

Pressure

logP > 3.0; VP

< 2 mmHg

2

Model 2

logP, TPSA

logP

TPSA

logP > 3.0;

TPSA < 60 Ų

The multiple linear regression (MLR) model built on five key physicochemical

descriptors

—

boiling point, logP, vapor pressure, TPSA, and molecular weight

—

produced a

statistically significant prediction of algicidal activity (

R²

= 0.68,

p

< 0.001). The final QSAR

equation was:

Algicidal Activity Score =+ 0.021 × logP − 0.030 × Vapor Pressure + 0.015 × Boiling

Point − 0.018 × TPSA − 0.005 × MW + 1.05

Among the predictors, logP and boiling point were positively associated with activity (

p

<

0.01), whereas vapor pressure and TPSA showed negative influence. Molecular weight had a
minor effect.

The model was validated using 5-fold cross-validation and yielded strong performance

metrics:

2025

JUNE

NEW RENAISSANCE

INTERNATIONAL SCIENTIFIC AND PRACTICAL CONFERENCE

VOLUME 2

|

ISSUE 6

450

•

Cross-

validated R² (Q²)

= 0.63

•

RMSE = 0.66

•

MAE = 0.54

Residual plots showed no signs of heteroscedasticity or systematic bias. The predicted vs.

observed activity scatterplot showed tight clustering along the identity line (y = x), confirming
good predictive reliability.

Leverage plots and standardized residual analysis indicated that 93% of compounds fell

within the model's applicability domain, suggesting strong generalizability across structurally
diverse molecules. (Figure 1 and 2 )

Figure 1: 2D Structural Formula of Benzothiazole:

Figure 2: 3D Structural Formula of Benzothiazole:

Discussion:

The structure

–

activity relationship (SAR) analysis clearly highlighted the

influence of specific structural features on the algicidal potency of organic compounds.
Halogenated aromatic rings, such as chloro- and bromo-substituted phenyl groups, were strongly
associated with enhanced activity, most likely due to increased lipophilicity and membrane
permeability. Additionally, heterocyclic systems like pyrroles and thiazoles were found to exhibit
greater efficacy compared to their aliphatic analogs, reinforcing the role of aromatic electron
density and ring strain in bioactivity. In contrast, the presence of polar substituents such as
hydroxyl and carboxyl groups was linked to reduced activity, likely due to their limited ability to
penetrate lipid-rich algal membranes.

2025

JUNE

NEW RENAISSANCE

INTERNATIONAL SCIENTIFIC AND PRACTICAL CONFERENCE

VOLUME 2

|

ISSUE 6

451

The hierarchical clustering analysis grouped compounds into high, moderate, and low

activity categories, enabling a qualitative framework for descriptor-based modeling. These
structural classifications directly informed the selection of physicochemical descriptors for
quantitative SAR (QSAR) analysis.

Two regression-based QSAR models were constructed using physicochemical parameters.

Model 1, based on logP and vapor pressure, demonstrated that higher lipophilicity and lower
volatility correlated strongly with greater algicidal activity. This finding aligns with earlier
observations in pesticide research, where compound retention and membrane affinity significantly
influenced efficacy. Model 2, incorporating logP and topological polar surface area (TPSA),
showed that higher polarity negatively impacted bioactivity, confirming that reduced TPSA
improves membrane diffusion and interaction with algal cell targets.

Both models were statistically robust, with Model 1 yielding an R² of 0.68 and Model 2

offering similar predictive accuracy. Cross-validation metrics, including RMSE (0.66) and MAE
(0.54), indicated reliable performance. The predicted versus observed activity plots and leverage
analysis further confirmed model stability and generalizability, with 93% of the compounds falling
within the applicability domain.

The 3D molecular visualization of benzothiazole provided additional insight into spatial

geometry and bond lengths, reinforcing the structural symmetry and aromatic character that likely
contribute to its bioactivity. The combination of 2D and 3D representations, along with descriptor-
based modeling, provides a holistic framework for predicting algicidal efficacy without the need
for extensive in vitro screening.

Overall, this study demonstrates the utility of combining SAR logic with QSAR regression

techniques to identify key molecular determinants of algicidal activity. These findings pave the
way for the design of more efficient and environmentally responsible algicides through descriptor-
driven pre-screening.

Conclusion:

This study successfully demonstrated that the algicidal activity of organic

compounds can be effectively predicted using structure

–

activity relationship (SAR) and

quantitative structure

–

activity relationship (QSAR) models based on key physicochemical

descriptors. Through qualitative SAR analysis, it was shown that halogenated aromatic rings and
heterocyclic systems are positively associated with higher bioactivity, while polar substituents tend
to reduce efficacy.

The developed QSAR models

—

especially those incorporating descriptors such as logP,

vapor pressure, and topological polar surface area (TPSA)

—

exhibited strong predictive power and

statistical reliability. Model 1, in particular, highlighted the significance of lipophilicity and
volatility as primary factors influencing algicidal potency. Both models demonstrated good cross-
validation performance and a wide applicability domain, encompassing over 90% of the analyzed
compounds.

Moreover, 3D molecular visualization provided deeper insight into the spatial orientation

of atoms and bond lengths, supporting the mechanistic interpretation of descriptor-based activity
trends.

Taken together, the findings of this research confirm the viability of descriptor-driven pre-

screening for environmentally safer and more efficient algicidal agents. Future studies should

2025

JUNE

NEW RENAISSANCE

INTERNATIONAL SCIENTIFIC AND PRACTICAL CONFERENCE

VOLUME 2

|

ISSUE 6

452

expand the descriptor set, integrate machine learning approaches, and validate these models
against broader algal species and environmental conditions to further enhance their applicability.

REFERENCES

1.

Шамшурин, А. А., & Кример, М. З.

(1976).

Физико

-

химические свойства

пестицидов: Справочник

(с. 3

-

11). Москва: Химия.

2.

Zhao, J., Xu, Y., & Liu, Y. (2018). Physicochemical property-based modeling of
bioactivity in pesticide research: Recent advances and applications.

Pest Management

Science

, 74(5), 1032

–

1044.

3.

Tornero-Velez, R., Davis, J. A., & Hines, R. N. (2012). Physicochemical properties and
environmental fate of pesticides. In R. Krieger (Ed.),

Hayes' Handbook of Pesticide

Toxicology

(3rd ed., pp. 145

–

168). Academic Press.

4.

Xayrullo o‘g, P. U. B. (2025, June). CHEMICAL ANALYSIS

-BASED ASSESSMENT

OF THE HERBICIDAL EFFICIENCY OF AZIDO-SUBSTITUTED TRIAZINES.
In

CONFERENCE OF ADVANCE SCIENCE & EMERGING TECHNOLOGIES

(Vol. 1,

No. 2, pp. 53-62).

5.

Zhang, W., He, M., & Chen, C. (2020). QSAR analysis and prediction of acaricidal activity
of novel organic compounds using multiple linear regression.

Journal of Agricultural and

Food Chemistry

, 68(12), 3412

–

3419.

6.

Maxsudjon T. et al. SYNTHESIS AND STUDY OF MIXED-LIGAND COMPLEX
COMPOUNDS BASED ON ALANINE AND 3D-METAL BENZOATES //Universum:

химия

и

биология

.

–

2022.

–

№.

6-4 (96).

–

С.

17-21.

7.

Xayrullo o'g P. U. et al. Using natural plant extracts as acid-base indicators and pKa value
calculation method //fan va ta'lim integratsiyasi (integration of science and education).

–

2024.

–

Т

. 1.

–

№.

3.

–

С. 80

-85.

8.

Xayrullo o'g P. U. et al. The essence of the research of synthesis of natural indicators,
studying their composition and dividing them into classes //fan va ta'lim integratsiyasi
(integration of science and education).

–

2024.

–

Т

. 1.

–

№.

3.

–

С. 50

-55.

9.

Nurmonova E., Berdimuratova B., Pardayev U. DAVRIY SISTEMANING III A GURUHI

ELEMENTI ALYUMINIYNING DAVRIY SISTEMADA TUTGAN O ‘RNI VA FIZIK

-

KIMYOVIY XOSSALARINI TADQIQ ETISH //Modern Science and Research.

–

2024.

–

Т

. 3.

–

№. 10. –

С

. 517-526.

10.

Kim, H. J., Park, H. M., & Lee, S. H. (2019). Structure

–

activity relationships of plant-

derived miticides and their physicochemical descriptors.

Journal of Pest Science

, 92(4),

1435

–

1446.

11.

Хайдаров Г. Ш. и др. СИНТЕЗ И БИОЛОГИЧЕСКАЯ АКТИВНОСТЬ
ГИДРОХЛОРИД ХИНАЗОЛИН

-4-

ОНА //“Fan va taʼlim integratsiyasi” jurnalining

Tahrir hay’ati tarkibi.

12.

Yap, C. W. (2011). PaDEL-Descriptor: An open source software to calculate molecular
descriptors and fingerprints.

Journal of Computational Chemistry

, 32(7), 1466

–

1474.

13.

Xayrullo o‘g, P. U. B. (2025). INVESTIGATION OF THE REPELLENT ACTIVITY

AGAINST

IXODID

TICKS

BASED

ON

THE

STRUCTURAL

AND

2025

JUNE

NEW RENAISSANCE

INTERNATIONAL SCIENTIFIC AND PRACTICAL CONFERENCE

VOLUME 2

|

ISSUE 6

453

PHYSICOCHEMICAL PROPERTIES OF DIBUTYL ADIPATE.

TANQIDIY NAZAR,

TAHLILIY TAFAKKUR VA INNOVATSION G ‘OYALAR

,

2

(1), 265-273.

14.

Jiemuratova A., Pardayev U., Bobojonov J. COORDINATION INTERACTION
BETWEEN ANTHRANILIC LIGAND AND D-ELEMENT SALTS DURING
CRYSTAL FORMATION: A STRUCTURAL AND SPECTROSCOPIC APPROACH
//Modern Science and Research.

–

2025.

–

Т

. 4.

–

№.

5.

–

С. 199

-201.

15.

Xoliyorova S., Tilyabov M., Pardayev U. Explaining the basic concepts of chemistry to
7th grade students in general schools based on steam //Modern Science and Research.

–

2024.

–

Т

. 3.

–

№. 2. –

С

. 362-365.

16.

Xayrullo o'g P. U. B., Rajabboyovna K. X. Incorporating Real-World Applications into
Chemistry Curriculum: Enhancing Relevance and Student Engagement //FAN VA
TA'LIM INTEGRATSIYASI (INTEGRATION OF SCIENCE AND EDUCATION).

–

2024.

–

Т

. 1.

–

№.

3.

–

С. 44

-49.

17.

Xayrullo o'g P. U. B., Umurzokovich T. M. Inquiry-Based Learning in Chemistry
Education: Exploring its Effectiveness and Implementation Strategies //FAN VA TA'LIM
INTEGRATSIYASI (INTEGRATION OF SCIENCE AND EDUCATION).

–

2024.

–

Т

.

1.

–

№.

3.

–

С. 74

-79.

18.

Cherkasov, A., Muratov, E. N., Fourches, D., et al. (2014). QSAR modeling: where have
you been? Where are you going to?

Journal of Medicinal Chemistry

, 57(12), 4977

–

5010.

19.

Pardayev U. et al. THE EFFECTS OF ORGANIZING CHEMISTRY LESSONS BASED
ON THE FINNISH EDUCATIONAL SYSTEM IN GENERAL SCHOOLS OF
UZBEKISTAN //Journal of universal science research.

–

2024.

–

Т

. 2.

–

№.

4.

–

С. 70

-74.

20.

Yang, Z., Wang, X., & Xu, Y. (2019). QSAR study on algicidal activity of substituted
phenols.

Chemosphere

, 231, 499

–

505.

21.

Choriqulova D. et al. The role of the method of teaching chemistry to students using the"
assessment" method //Modern Science and Research.

–

2024.

–

Т

. 3.

–

№.

11.

–

С. 256

-

264.\

22.

Zhao, Y., Zhou, Y., & Wang, H.

(2017). Physicochemical basis of algicidal activity of

organic molecules.

Ecotoxicology and Environmental Safety

, 139, 209

–

215.